On 12/7/05 12:49 PM, "Tim Smith" <tim_smith_666 at="" yahoo.com=""> wrote:
> Hi all,
>
> A newbie to Bioconductor, so please forgive the naive question..
>
> I had two sets of data (normalized affy data). The first set
contains
> cancerous expression data for 40 patients, and the second set
contains
> expression data for 30 'normal' (i.e non cancerous) patients.
>
> I wanted to find out the genes that are 'differentially expressed'
in the
> two sets. Which package, and which functions can I use? Any
reasonably popular
> method/package would work for me.
Welcome to bioconductor.
See here for a full list. Best to do a little reading, as each method
has
potentially different assumptions, etc.
http://www.bioconductor.org/packages/bioc/1.7/src/contrib/html/
That said, look at siggenes which contains the SAM method; the
multtest
package for traditional t-tests; and the limma package.
And just to be sure, when you say you have two sets of normalized
data, you
mean that they were normalized together or separately?
Sean
On Dec 7, 2005, at 3:17 PM, Tim Smith wrote:
> Thanks for the replies Sean and Naomi. I feel I'm on the right
track.
>
> I did the following:
>
>
> sam.out <- sam(zz, ncl, method = "d.stat", delta = NULL,
> n.delta = 10, p0 = NA,
> lambda = seq(0, 0.95, 0.05), ncs.value = "max", ncs.weights =
NULL,
> gene.names = dimnames(zz)[[1]])
>
>
> sam.sum <-summary(sam.out,2,ll=FALSE)
>
>
> Now, I wanted to get to the d-value matrix. The documentation
> says that I can get this by the line:
>
> # The matrix containing the d-values, q-values etc. of the
> # differentially expressed genes can be obtained by
>
>
> sam.outATmat.sig # sam.out at mat.sig
>
> but when I do this, I get :
>
> Error: no slot of name "mat.sig" for this object of class "SAM"
>
> How can I access the matrix to get at the underlying d-values
> etc..??
>
> many thanks.
Without going into too much detail, in R you have S3 objects and S4
objects. You use $ for S3 objects and @ for S4 objects. Valid names
may be found by names(object) for S3 objects and slotNames(object)
for S4 objects.
Kasper
> Sean Davis <sdavis2 at="" mail.nih.gov=""> wrote:
>
>
>
>
> On 12/7/05 12:49 PM, "Tim Smith" wrote:
>
>> Hi all,
>>
>> A newbie to Bioconductor, so please forgive the naive question..
>>
>> I had two sets of data (normalized affy data). The first set
contains
>> cancerous expression data for 40 patients, and the second set
>> contains
>> expression data for 30 'normal' (i.e non cancerous) patients.
>>
>> I wanted to find out the genes that are 'differentially expressed'
>> in the
>> two sets. Which package, and which functions can I use? Any
>> reasonably popular
>> method/package would work for me.
>
> Welcome to bioconductor.
>
> See here for a full list. Best to do a little reading, as each
> method has
> potentially different assumptions, etc.
>
> http://www.bioconductor.org/packages/bioc/1.7/src/contrib/html/
>
> That said, look at siggenes which contains the SAM method; the
> multtest
> package for traditional t-tests; and the limma package.
>
> And just to be sure, when you say you have two sets of normalized
> data, you
> mean that they were normalized together or separately?
>
> Sean
>
>
>
>
>
>
> ---------------------------------
>
>
> [[alternative HTML version deleted]]
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
I have not used this particular Bioconductor package. But I have
learned in general how to deal with the various types of objects that
the Bioconductor developers have created for us.
class(myobject)
tells you the class of "myobject"
?myclass
should give the documentation of what the class "myclass" consists of
method(myobject)
tells you what methods (and slots) work with "myobject"
names(myobject)
tells you what the components of "myobject" are.
Using these functions and the documentation, (vignettes and html
help) I can usually figure out what I need. If not ... there is
always the mailing list.
--Naomi
At 06:17 PM 12/7/2005, Tim Smith wrote:
>Thanks for the replies Sean and Naomi. I feel I'm on the right track.
>
> I did the following:
>
>
> sam.out <- sam(zz, ncl, method = "d.stat", delta = NULL,
> n.delta = 10, p0 = NA,
> lambda = seq(0, 0.95, 0.05), ncs.value = "max", ncs.weights =
NULL,
> gene.names = dimnames(zz)[[1]])
>
>
> sam.sum <-summary(sam.out,2,ll=FALSE)
>
>
> Now, I wanted to get to the d-value matrix. The documentation
> says that I can get this by the line:
>
> # The matrix containing the d-values, q-values etc. of the
> # differentially expressed genes can be obtained by
>
>
> sam.outATmat.sig # sam.out at mat.sig
>
> but when I do this, I get :
>
> Error: no slot of name "mat.sig" for this object of class "SAM"
>
> How can I access the matrix to get at the underlying d-values
etc..??
>
> many thanks.
>
>Sean Davis <sdavis2 at="" mail.nih.gov=""> wrote:
>
>
>
>
>On 12/7/05 12:49 PM, "Tim Smith" wrote:
>
> > Hi all,
> >
> > A newbie to Bioconductor, so please forgive the naive question..
> >
> > I had two sets of data (normalized affy data). The first set
contains
> > cancerous expression data for 40 patients, and the second set
contains
> > expression data for 30 'normal' (i.e non cancerous) patients.
> >
> > I wanted to find out the genes that are 'differentially expressed'
in the
> > two sets. Which package, and which functions can I use? Any
> reasonably popular
> > method/package would work for me.
>
>Welcome to bioconductor.
>
>See here for a full list. Best to do a little reading, as each method
has
>potentially different assumptions, etc.
>
>http://www.bioconductor.org/packages/bioc/1.7/src/contrib/html/
>
>That said, look at siggenes which contains the SAM method; the
multtest
>package for traditional t-tests; and the limma package.
>
>And just to be sure, when you say you have two sets of normalized
data, you
>mean that they were normalized together or separately?
>
>Sean
>
>
>
>
>
>
>---------------------------------
>
>
> [[alternative HTML version deleted]]
>
>_______________________________________________
>Bioconductor mailing list
>Bioconductor at stat.math.ethz.ch
>https://stat.ethz.ch/mailman/listinfo/bioconductor
Naomi S. Altman 814-865-3791 (voice)
Associate Professor
Dept. of Statistics 814-863-7114 (fax)
Penn State University 814-865-1348
(Statistics)
University Park, PA 16802-2111
This is a bug in the example section of the help files for the
function sam.
It should actually be
# The matrix containing the d-values, q-values etc. of the
# differentially expressed genes can be obtained by
sum.sam3 at mat.sig
and not sam.out at mat.sig. So in your case,
> sam.sum at mat.sig
will give you the above matrix.
Sorry for this bug. I will fix it pretty soon. You can find a correct
(and
tested) description in the vignette of siggenes. See
> vignette("siggenes")
or if you use Windows click on Vignettes --> siggenes --> siggenes.
Best,
Holger
> --- Urspr?ngliche Nachricht ---
> Von: Tim Smith <tim_smith_666 at="" yahoo.com="">
> An: Sean Davis <sdavis2 at="" mail.nih.gov="">, Bioconductor
> <bioconductor at="" stat.math.ethz.ch="">
> Betreff: Re: [BioC] Evaluating differentiall expressed genes
> Datum: Wed, 7 Dec 2005 15:17:34 -0800 (PST)
>
> Thanks for the replies Sean and Naomi. I feel I'm on the right
track.
>
> I did the following:
>
>
> sam.out <- sam(zz, ncl, method = "d.stat", delta = NULL, n.delta
= 10,
> p0 = NA,
> lambda = seq(0, 0.95, 0.05), ncs.value = "max", ncs.weights =
NULL,
> gene.names = dimnames(zz)[[1]])
>
>
> sam.sum <-summary(sam.out,2,ll=FALSE)
>
>
> Now, I wanted to get to the d-value matrix. The documentation says
that
> I can get this by the line:
>
> # The matrix containing the d-values, q-values etc. of the
> # differentially expressed genes can be obtained by
>
>
> sam.outATmat.sig # sam.out at mat.sig
>
> but when I do this, I get :
>
> Error: no slot of name "mat.sig" for this object of class "SAM"
>
> How can I access the matrix to get at the underlying d-values
etc..??
>
> many thanks.
>
> Sean Davis <sdavis2 at="" mail.nih.gov=""> wrote:
>
>
>
>
> On 12/7/05 12:49 PM, "Tim Smith" wrote:
>
> > Hi all,
> >
> > A newbie to Bioconductor, so please forgive the naive question..
> >
> > I had two sets of data (normalized affy data). The first set
contains
> > cancerous expression data for 40 patients, and the second set
contains
> > expression data for 30 'normal' (i.e non cancerous) patients.
> >
> > I wanted to find out the genes that are 'differentially expressed'
in
> the
> > two sets. Which package, and which functions can I use? Any
reasonably
> popular
> > method/package would work for me.
>
> Welcome to bioconductor.
>
> See here for a full list. Best to do a little reading, as each
method has
> potentially different assumptions, etc.
>
> http://www.bioconductor.org/packages/bioc/1.7/src/contrib/html/
>
> That said, look at siggenes which contains the SAM method; the
multtest
> package for traditional t-tests; and the limma package.
>
> And just to be sure, when you say you have two sets of normalized
data,
> you
> mean that they were normalized together or separately?
>
> Sean
>
>
>
>
>
>
> ---------------------------------
>
>
> [[alternative HTML version deleted]]
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
>
--
On 7 Dec 2005, naomi at stat.psu.edu wrote:
> I have not used this particular Bioconductor package. But I have
> learned in general how to deal with the various types of objects
> that the Bioconductor developers have created for us.
These are good suggestions and I use this approach all the time to get
a feel for code I'm not familiar with.
> class(myobject)
>
> tells you the class of "myobject"
>
> ?myclass
>
> should give the documentation of what the class "myclass" consists
> of
For S4 classes, it will sometimes be necessary to use:
class ? myclass
Another thing to do is to scan the index of functions:
help(package = "mypackage")
+ seth
While we are on the topic of strategies for learning Bioconductor, I
would like to recommend the wonderful R tutorial that is available on
the "html help" page.
"An Introduction to R"
I have been using some version or other of R (or S) since 1982, and I
still learned a lot from this document. I also use it to teach a
course that covers R for complete novices.
--Naomi
At 09:45 AM 12/9/2005, Seth Falcon wrote:
>On 7 Dec 2005, naomi at stat.psu.edu wrote:
> > I have not used this particular Bioconductor package. But I have
> > learned in general how to deal with the various types of objects
> > that the Bioconductor developers have created for us.
>
>These are good suggestions and I use this approach all the time to
get
>a feel for code I'm not familiar with.
>
> > class(myobject)
> >
> > tells you the class of "myobject"
> >
> > ?myclass
> >
> > should give the documentation of what the class "myclass" consists
> > of
>
>For S4 classes, it will sometimes be necessary to use:
>
>class ? myclass
>
>Another thing to do is to scan the index of functions:
>
>help(package = "mypackage")
>
>+ seth
>
>_______________________________________________
>Bioconductor mailing list
>Bioconductor at stat.math.ethz.ch
>https://stat.ethz.ch/mailman/listinfo/bioconductor
Naomi S. Altman 814-865-3791 (voice)
Associate Professor
Dept. of Statistics 814-863-7114 (fax)
Penn State University 814-865-1348
(Statistics)
University Park, PA 16802-2111